Identification of Differentially Expressed Genes ...

7 downloads 178584 Views 1MB Size Report
Aug 17, 2016 - and “Honeycrisp” apple endocarp and peel tissues. ..... In order to account for this discrepancy, all predicted matches were included resulting.
horticulturae Article

Identification of Differentially Expressed Genes between “Honeycrisp” and “Golden Delicious” Apple Fruit Tissues Reveal Candidates for Crop Improvement Scott Schaeffer 1,2,† , Christopher Hendrickson 2,‡ , Rachel Fox 2 and Amit Dhingra 1,2, * 1 2

* † ‡

Molecular Plant Science Program, Washington State University, Pullman, WA 99164, USA; [email protected] Department of Horticulture, Washington State University, Pullman, WA 99164, USA; [email protected] (C.H.); [email protected] (R.F.) Correspondence: [email protected]; Tel.: +1-509-335-3625 Current address: Department of Agriculture-Agriculture Research Service, Children’s Nutrition Research Center, Baylor College of Medicine, Houston, TX 77030, USA. Current address: Department of Mathematics and Natural Sciences, National University, La Jolla, CA 92037, USA.

Academic Editors: Lijun Wang and Douglas D. Archbold Received: 1 February 2016; Accepted: 6 August 2016; Published: 17 August 2016

Abstract: Cultivars of the same species exhibit a large degree of variation in fruit quality traits, which can be directly influenced by differences in gene expression due to allelic variations and interactions with the environment. For Malus × domestica Borkh. (apple), fruit quality traits, including color, texture, aroma, flavor profile, and shelf life, are of utmost economic importance. In order to identify genes potentially influencing these traits, a direct comparative transcriptome profiling approach, based on the differential display technique, was performed using “Golden Delicious” and “Honeycrisp” apple endocarp and peel tissues. A total of 45 differentially expressed sequence tags were identified between the two apple varieties. Reanalysis of a previously published fruit developmental microarray expression experiment revealed that only one of the 45 sequence tags was represented on the array. Differential expression of 31 sequence tags from the peel tissue was validated using quantitative reverse transcription PCR, confirming the robustness of the differential display approach to quickly identify differentially expressed sequence tags. Among these were genes annotated to be involved in ripening, phytohormone signaling, transcription factors, and fruit texture. This work demonstrates yet again the utility of the differential display technique to rapidly identify genes related to desirable traits. Keywords: differential display; differential expression; fruit quality; Malus × domestica Borkh

1. Introduction Fruit quality traits vary across different cultivars of the same species during development, storage, and/or transit to market. In pome fruit, these traits can be broadly classified as peel or endocarp-specific. Peel-specific traits include fruit color, photo-protective capacity, scalding susceptibility, lenticel injury, aroma, etc. Endocarp-specific traits include propensity for rotting and mold development, hypoxic injury, flesh-sweetness, acidity, and textural qualities. Adding further complexity, these traits are under polygenic control and can vary by season, soil conditions and elevation [1–3]. Development of new tree fruit cultivars remains a laborious and time-consuming process. Juvenility periods and a perennial fruit-bearing nature can require at least five years to study the Horticulturae 2016, 2, 11; doi:10.3390/horticulturae2030011

www.mdpi.com/journal/horticulturae

Horticulturae 2016, 2, 11

2 of 17

effect of genes in progeny of a controlled cross. Additionally, desirable traits in tree fruits may be highly dependent on genetic and environmental interactions during the course of fruit development and maturation, which can complicate attempts at performing uniform phenotyping. As a result, identification of causal genetic mechanisms underlying traits of interest in tree fruits can be difficult. Additionally, genomic structure of such species can compound challenges. Theorized duplication events have yielded allotetraploid genome structures in several perennial tree fruits including apple and pear [4,5]. This can result in numerous functionally redundant or interrelated families of genes involved in controlling traits of interest in tree fruit. To facilitate crop improvement in such systems, desirable genes must first be functionally validated prior to them being selectively introgressed into future genetic selections. Variation in fruit quality traits can be influenced by numerous genetic factors including, but not limited to, allelic variation, gene copy number, and transcript variants. Differential gene and allelic expression may also result in aberrant folding or post-translational modification of proteins within the members of the same species [6,7]. While the causes of the phenotypic differences can vary, approaches used to identify causal genetic elements have been largely uniform. To identify the genetic elements, researchers rely upon segregation probabilities of DNA polymorphisms in populations segregating for the desired traits with the use of molecular markers such as RFLPs (Restriction Fragment Length Polymorphism), SNPs (Single Nucleotide Polymorphism), and microsatellites [8,9]. Environmental cues, polygenic control of quantitative traits, and allelic expression differences can complicate the use of molecular markers in identification of underlying genetic elements in an apple. For these reasons, the exclusive use of marker-based approaches to identify underpinnings of desirable fruit phenotypes can be challenging. While not feasible for all traits, a gene-linked mechanistic understanding of desirable fruit traits wherever possible may provide a reliable and reproducible resource for crop improvement. The Malus research community has developed and deployed molecular markers from densely populated maps of all Malus × domestica Borkh. linkage groups. Fruit firmness in apple was reported to be linked with the allelic composition of aminocyclopropane-1-carboxylic acid synthase (ACS) and oxidase (ACO) genes that regulate ethylene biosynthesis in climacteric fruit [10,11]. Zhu and Barrett [12] applied this information to nearly 100 apple lines and concluded that those homozygous for ACS1-2 and ACO1-1 had firmer fruit both at harvest and after typical commercial postharvest storage, compared to other genotypes. Additional correlations have been made between a Malus SSR marker and resistance to Erwinia amylovra (fireblight) [13]. These signatures of economically-important traits can be used to develop improved germplasm for the commercial market. Recently, a wealth of additional tree fruit genomic resources have been established, with the release of the genomes from the “Golden Delicious” apple, sweet orange, peach, and Chinese white pear, as well as the recently announced European pear genomes [4,5,14–16]. These resources are enabling the understanding of genetic causes underlying numerous traits in tree fruit such as resistance to pathogens, fruit sweetness, acidity, texture, peel and flesh coloring, and responses to numerous postharvest storage regimes. One of the less used approaches to identify the genetic basis for phenotypic differences between two varieties is to perform a direct comparison of gene expression in a spatio-temporal context. One such small scale approach for comparative transcriptome profiling is the differential display reverse transcriptase polymerase chain reaction, or DD [17]. The DD technique can amplify all poly-adenylated transcripts allowing for direct capture of actively expressed genes and alleles in a comparative context. The resulting gene fragment amplicons can be isolated based upon either their presence or absence, or based on differential intensity on a polyacrylamide gel that is indicative of differential expression. Identified fragments are cloned, sequenced, and annotated to infer their biological function. This approach has been utilized in many plant systems to identify putative targets involved in many traits or processes including reproductive roles in semigametic Pima cotton (Gossypium barbadense), heat responsive genes in peach (Prunus persica), hormone and abiotic stress changes in the gene expression of tea (Camelina sinensis), and ripening-related genes in banana

Horticulturae 2016, 2, 11

3 of 17

(Musa acuminata AAA) [18–21]. In each of these instances, DD aided in direct identification of genes potentially involved in the regulation of the phenotypes being studied. Despite the labor and time-related challenges associated with performing a DD approach, the technique retains some significant advantages over newer next-generation sequencing based approaches. Differential display offers an inexpensive means of identifying candidate genes without the complexities associated with RNA sequencing (RNA-seq) based approaches. The presence or absence of a PCR amplicon in a DD polyacrylamide gel provides a positive result that does not require a reference genome and advanced computational resources associated with comparative RNA-seq workflows. Further, DD can aid in the identification of potentially important regulatory genes between samples which can be overlooked using microarray-based comparisons. The complex data capture, assembly and statistical analysis associated with microarray and next-generation sequencing approaches can result in missing or misidentifying the genes that control traits of interest. Caution is prescribed in carefully examining novel transcripts identified via RNA-seq before proceeding to biological experiments. Available apple germplasm abounds in diversity of fruit quality traits. The apple cultivar “Honeycrisp” (HC) falls into a category of “JFC high quality” indicating a juicy and crispy-textured flesh, while “Golden Delicious” (GD) is classified as “American/European dessert”. HC is typified by a color profile which is 60% orange/red on top of a yellow base color, large size, as well as having an exceptional texture and juiciness [22]. GD, on the other hand, has a completely yellow color and texture that is juicy, firm and crisp. HC apples maintain a crisp texture up to 6 months after harvest even without controlled atmosphere conditions due to maintenance of cell wall integrity and cellular turgor potential [23]. These two apple varieties are popular among consumers, but exhibit highly diverse fruit quality characteristics through ontogeny and ripening. In this study, we investigated the differences in the expression of peel and flesh specific genes of GD and HC apples at two different developmental stages. Results obtained from this work identified several important genes related to fruit quality and demonstrate the utility of an inexpensive method for direct identification of candidate genes that could be utilized for development of improved apple varieties. 2. Materials and Methods 2.1. Sampling Strategy HC and GD apples were harvested during two growing seasons at 125 or 129 and 150 or 160 days after anthesis (DAA), respectively. The time points sampled in this study were chosen due to their correlation with early and late stages of the ripening process in an apple, which typically begins around 95 days after anthesis (DAA) and concludes at around 150 DAA [1]. Additionally, these samples represent apple fruit at pre-climacteric and climacteric stages of maturity [2]. Tissue processing was performed directly in the orchard where the peel and endocarp samples were obtained from nine apples derived from three trees of the same variety. Samples were immediately frozen in liquid nitrogen. Tissues were stored at −80 ◦ C and subsequently ground for three cycles of four minutes at top speed using a SPEX SamplePrep 6870 Freezer mill (Metuchen, NJ, USA). Total RNA was extracted from ground tissue using Plant RNeasy Minikit according to the manufacturer’s instructions (Qiagen, Valencia, CA, USA). RNA was quantified using a Nanodrop 8000 spectrophotometer (Wilmington, DE, USA) and its integrity was confirmed with a denaturing formaldehyde MOPS gel. 2.2. Differential Display Differential displays were performed on HC endocarp, HC peel, GC endocarp and GC peel samples. Removal of DNA contamination from RNA was performed with the MessageClean Kit (GenHunter, Nashville, TN, USA) according to manufacturer’s instructions. Synthesis of cDNA and PCR amplification was performed using the GenHunter RNAimage Kit (GenHunter) according to the manufacturer’s instructions. Primers 193 to 200 (GenHunter) in combination with the anchoring

Horticulturae 2016, 2, 11

4 of 17

poly-T, H-T11 A, were used for PCR amplification. 33 P-labeled cDNA fragments were separated on a 6% acrylamide denaturing gel. After imaging through radiography, bands were excised if differential expression was observed between HC and GD tissues. 2.3. Cloning and Sequencing Excised bands were boiled for 10 min with 40 µL of deionized water in a 1.5 mL microcentrifuge tube. Re-amplification was performed using 1 µL of the recovered DNA as a template and the differential display primers used for their initial amplification. PCR amplicons were cleaned using the QIAquick PCR Purification Kit (Qiagen) and subsequently ligated into the pGEM-T Easy vector (Promega, Madison, WI, USA) according to the manufacturer’s protocol. Ligated fragments were then transformed into Escherichia coli XL1-Blue cells via the heat-shock method. Colony PCR was used to confirm the presence of an insert using M13 forward and reverse primers. Plasmids were sequenced using Sanger sequencing with M13 forward primers at Eurofins MWG Operon. Gene IDs as predicted from the apple genome of the sequenced fragments are provided in Table S1. 2.4. Annotation of Differential Display Fragments BLASTN was performed for all sequenced differential display fragments against the predicted gene sequences from the apple genome release 1.0. As differential display sequences were derived from the 3’ end of transcripts and could contain untranslated regions, a second BLASTN was performed against the NCBI EST database for Rosaceae sequences (NCBI taxid: 3745). Matching EST sequences were subsequently mapped to full length apple gene sequences with BLASTN which were extracted for subsequent analysis. All full length apple gene sequences were then analyzed by Blast2GO against the NCBI nr database to determine putative functions [3,4]. Localization of each protein sequence was performed with TargetP using plant networks with default parameters [5]. A final BLASTX step was performed against the NCBI protein collections from Arabidopsis thaliana and Solanum lycopersicum to identify homologs in these systems, with the top result from each species with an E-value of at least 1E−20 listed. The description, if more detailed than the original Blast2GO hit, was also retained. 2.5. Quantitative Real Time PCR Validation of Differential Expression RNA from HC and GD peel tissue collected over two growing seasons was extracted as described above. Validation of differential expression using qPCR was performed with peel tissues. Samples were DNAse-treated using the DNA-free kit (Ambion) according to the manufacturer’s instructions. RNA was quantified using Nanodrop 8000 and its integrity was tested on a 1.0% agarose gel. First-strand cDNA was synthesized using the SuperScript VILO cDNA Synthesis Kit (Life Technologies, Carlsbad, CA, USA) following manufacturer’s instructions. Concentration was diluted to 50 ng/µL with deionized water. Validation of differential expression was performed for 31 apple genes: MDP0000200646, MDP0000037251, MDP0000875654, MDP0000296716, MDP0000128468, MDP0000712586, MDP0000618650, MDP0000920333, MDP0000152947, MDP0000883284, MDP0000253074, MDP0000213808, MDP0000547450, MDP0000310811, MDP0000232309, MDP0000161275, MDP0000176723, MDP0000304285, MDP0000172863, MDP0000180389, MDP0000523205, MDP0000138340, MDP0000166116, MDP0000220601, MDP0000237908, MDP0000273484, MDP0000286959, MDP0000292888, MDP0000320533, MDP00005772242, and MDP0000697474. Primers for PCR were designed with the assistance of custom scripts based on predicted gene sequences from the apple genome project. These primers were then used in quantitative reverse-transcription PCR validation of DD results using iTaq Universal SYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA, USA) on a Stratagene MX3005P light cycler (Agilent, Santa Clara, CA, USA). All reactions were performed using 50 ng cDNA and carried out in triplicate along with a β-tubulin internal control. β-Tubulin was selected as it has been previously used as an internal control in several prior apple gene expression studies [6–8]. Efficiencies and Cq values were determined using the tool LinRegPCR [9]. Any reactions with an efficiency less than 1.8 or more than 2.2 were discarded from further analysis. Cq values were compared between the three replicates and,

Horticulturae 2016, 2, 11

5 of 17

in the case that the standard deviation exceeded one, one of the three outliers was omitted from the analysis. The Pfaffl correction was then applied to Cq values of remaining reactions. Fold change was determined by comparing the highest Cq value to a β-tubulin internal control [10]. To validate the identity of the genes, 10 random qPCR products were sequenced. These included MDP0000312808, MDP0000232309, MDP00000320533, MDP0000547450, MDP0000697474, MDP0000166116, MDP0000253074, MDP0000166116, MDP0000253074, MDP0000255887, MDP0000200646, and MDP0000920333. Products were sequenced using the forward and reverse primers that were originally used to amplify them. Sequences were assembled using Lasergene 11 SeqMan Pro (DNASTAR, Madison, WI, USA) and compared to the full length apple gene using a pairwise BLAST. 2.6. Comparison of Expression with Previous Studies Previously generated “Royal Gala” apple microarray expression data published in 2008 [1] was reanalyzed to identify data from genes recognized in the differential display experiment. All ESTs represented in the experiment were compared to the predicted apple gene sequences generated from the apple genome with BLASTN. The top hits with an e-value lower than 1E−10 were then assigned the apple gene identifiers. These identifiers were then compared with the genes represented in the differential display experiment. 3. Results and Discussion 3.1. Differential Display Apple ripening is characterized by the conversion of starches to simple sugars, decrease of total chlorophyll and photosynthetic activity, increase in total carotenoids, flesh softening and cell wall modification as well as accumulation of volatiles and flavor compounds [11–13]. Any differences in these properties would likely be mirrored in differences in gene expression. Differential display analysis of the peel tissues from two developmental stages of GD and HC tissues in apple, generated 115 bands corresponding to potentially differentially expressed genes. Of these, 105 fragments could be isolated which were then cloned into a plasmid and sequenced. The remaining ten could not be re-amplified after extraction from the PAGE (Polyacrylamide Gel Electrophoresis). BLASTN was performed for each sequence against the predicted gene set from the apple genome as well as the NCBI EST sequences available for Rosaceae (NCBI taxid: 3745). The apple genome and Rosaceae EST sequences similar to the 105 query sequences were extracted and processed through BLAST2GO workflow and the output is available in Table S2. For each matching EST sequence, a final BLAST was performed against the apple predicted gene set. This analysis resulted in the identification of 42 sequences from the differential display experiment, which had significant identity with a full-length apple gene. Three of these sequences, however, had different apple gene matches between the two BLAST methods. In order to account for this discrepancy, all predicted matches were included resulting in 45 sequences which were chosen for further characterization and validation. Blast2GO analysis was performed on matching full length apple sequences to predict potential functions associated with the sequences, while TargetP was used to predict the cellular localization of each protein (Table 1) [3–5,14]. Of these genes, seven are predicted to be secreted from the cell, while fewer are predicted to be localized to the mitochondria and the chloroplast. The most represented molecular functions in this group include transporter activity (GO:0005215, four sequences), sequence-specific DNA binding transcription factor activity (GO:0003700, three sequences), and protein binding (GO:0005515, three sequences). Relatively more information was available for the biological processes in which these proteins may participate, which include response to stress (GO:0006950, 11 sequences), response to biotic stimulus (GO:0009607, nine sequences), transport (GO:0006810, eight sequences), and catabolic process (GO:0009056, seven sequences). Molecular function and biological process gene ontologies for these genes are summarized in Figure 1. A final BLASTX was performed against the NCBI protein collections from Arabidopsis thaliana and Solanum lycopersicum to identify the homologs in these systems (Table 2).

Horticulturae 2016, 2, 11

6 of 17

Table 1. Identification of fragments isolated from an initial differential display procedure. Differential expression was assessed using “Honeycrisp” and “Golden Delicious” apple fruit peel and endocarp (HCP, HCE, GDP, and GCE, respectively) collected at 129 and 160 days after anthesis (DAA) in the 2007 growing season. Bands from differential display gels were excised, cloned and sequenced. Qualitative abundance was determined for each tissue and noted as either absent (−) present (+) or overexpressed (++). Sequences were matched to predicted apple genes via BLAST and analyzed for predicted localization using TargetP (M—mitochondria, C—chloroplast, S—secreted, _—no localization). Gene Identifier

Blast2GO Annotation

MDP0000037251 MDP0000128468 MDP0000129664 MDP0000138340 MDP0000152947 MDP0000161275 MDP0000166116 MDP0000172863 MDP0000176723 MDP0000180389 MDP0000200646 MDP0000213808 MDP0000220601 MDP0000232309 MDP0000233229 MDP0000234325 MDP0000237908 MDP0000253074 MDP0000255887 MDP0000273484 MDP0000281279 MDP0000286959 MDP0000292888 MDP0000296716 MDP0000304285

cinnamyl alcohol dehydrogenase-like protein abscisic acid stress ripening protein homolog 3-ketoacyl-thiolase NAC domain ipr003441 wound-induced protein mitochondrial substrate carrier family protein acyl:CoA ligase acetate-coa synthetase-like protein protein acyl:CoA ligase acetate-coa synthetase-like protein disease resistance protein at3g14460-like NAC domain ipr003441 probable ubiquitin conjugation factor e4-like zinc finger CCCH domain-containing protein 53-like transmembrane BAX inhibitor motif-containing protein 4 Unknown protein WWE protein-protein interaction domain family protein metallothionein-like protein abscisic acid stress ripening protein homolog TIR-NBS-LRR resistance protein SKP1-like protein Unknown protein dentin sialophosphoprotein GYF domain-containing protein ethylene-responsive transcription factor RAP2-7-like xanthine uracil permease family expressed

129 DAA-2007 Season

160 DAA-2007 Season

GDE

GDP

HCE

HCP

GDE

GDP

HCE

HCP

Cellular Localization

++ ++ ++ + + + ++ ++ + ++ + + + ++ ++ + ++ -

++ ++ ++ ++ + + ++ ++ + ++ + + + ++ + + + + + -

+ ++ + + + + + + + + + + + + + + + + ++ -

++ + + + + ++ + + + + + + + + + + + + ++ +

++ ++ ++ + + ++ ++ ++ + ++ + ++ + ++ ++ + ++ -

++ ++ ++ + + ++ + ++ ++ + ++ + ++ + ++ + + + + + -

+ ++ + + + + +++ + + ++ + + + + + + + + ++ -

++ + + + + ++ +++ + + ++ + ++ + ++ + + + + ++ +

_ M _ _ _ _ _ S _ S _ _ _ _ _ _ _ _ _ _ _ _ C _ _

Horticulturae 2016, 2, 11

7 of 17

Table 1. Cont. Gene Identifier

Blast2GO Annotation

MDP0000310811 MDP0000316244 MDP0000320533 MDP0000378585 MDP0000443265 MDP0000523205 MDP0000547450 MDP0000572242 MDP0000580900 MDP0000584042 MDP0000606453 MDP0000618650 MDP0000689999 MDP0000697474 MDP0000712586 MDP0000875654 MDP0000876817 MDP0000883284 MDP0000901731 MDP0000920333

cysteine protease inhibitor probable ADP-ribosylation factor GTPase-activating protein AGD15-like proteasome assembly chaperone at4g03420 f9h3_4 Unknown protein Unknown protein UNC93-like protein probable xyloglucan glycosyltransferase 12-like porin voltage-dependent anion-selective channel protein protein probable nitrite transporter at1g68570-like NAC domain ipr003441 protein disulfide isomerase reverse transcriptase protein SCAI-like hydroquinone glucosyltransferase Unknown protein PREDICTED: uncharacterized protein LOC100248602 [Vitis vinifera] Unknown protein putative protein phosphatase 2C 10

129 DAA-2007 Season

160 DAA-2007 Season

GDE

GDP

HCE

HCP

GDE

GDP

HCE

HCP

Cellular Localization

+ + + + + + ++ + ++ + -

+ + ++ ++ + + + + ++ + -

++ ++ + + + ++ + -

++ + + + + + + ++ + + +

++ + + + + + ++ ++ + + -

++ + + ++ ++ + + + ++ + + -

+ ++ + + + + + + + + -

+ + + + + + + + + + + +

M _ _ _ C _ _ _ _ S _ _ S _ S _ _ S S _

Horticulturae 2016, 2, 11

8 of 17

Table 2. Linking differentially expressed genes to changes in fruit traits. Functional annotation for differentially expressed genes was performed using Blast2GO, and additional BLASTX was performed against the proteomes of Arabidopsis thaliana and Solanum lycopersicum. Many of the differentially expressed genes influence important traits within the fruits. Gene ID

Location

Gene

MDP0000138340 MDP0000200646 MDP0000618650 MDP0000296716

chr1:12,793,865..12,795,081 chr1:12,796,087..12,797,315 chr15:14,098,214..14,102,470 chr3:9,357,237..9,360,502

NAC transcription factor-like 9 NAC transcription factor-like 9 NAC domain containing protein 75 AP2 transcription factor SlAP2d

MDP0000037251 MDP0000572242 MDP0000875654 MDP0000523205

chr1:6,887,281..6,890,898 chr15:1,231,253..1,235,097 chr7:16,340,110..16,341,528 chr7:5,077,703..5,080,814

Putative cinnamyl alcohol dehydrogenase 9 Probable xyloglucan glycosyltransferase 12-like Hydroquinone glucosyltransferase Flavonoid biosynthesis oxidoreductase protein

Sl Homolog

Potential Trait/Fruit Characteristic Influenced

Reference(s)

XP_004239709.1 XP_004239709.1 XP_004239596.1 NP_001234647.1

Stress response Stress response Stress response/general ripening regulation Ethylene-regulated response

[15] [15] [15] [16,17]

AT4G39330 AT4G07960 AT4G01070 AT4G10490

XP_004250169.1 XP_004238064.1 XP_004231207.1 NP_001233840.1

Fruit firmness Fruit firmness Volatile/polyphenol glycosylation Flavanoid synthesis

[18] [19] [20,21] [22]

At Homolog Transcriptional Regulation AT4G35580 AT4G35580 AT4G29230 AT2G28550

Secondary Biosynthetic Processes

Signaling

XP_004231630.1 XP_004237914.1 XP_004232186.1

Abiotic stress response, developmental response to sugar levels Abiotic stress response/ripening Ripening/abiotic stress response Ethylene production, development, ripening

[25–27] [28,29] [30]

No hit XP_004235041.1

Cell elongation/fruit size Apogamy

[31,32] [33]

No Hit XP_004249636.1 XP_004228480.1

Redox homeostasis/respiration Hypoxia-induced fermentation/respiration Abiotic stress tolerance

[34] [35,36] [37,38]

MDP0000128468

chr16:4,935,225..4,935,891

Abscisic acid stress ripening (ASR1) protein

No hit

NM_001247208.2

MDP0000176723 MDP0000920333 MDP0000213808

chr17:23,896,643..23,897,062 chr10:29,789,438..29,789,994 chr17:4,103,927..4,109,528

Acyl:CoA ligase acetate-CoA synthase-like protein Putative protein phosphatase 2C-10 putative ubiquitin conjugation factor E4

AT5G16370 AT1G34750 AT5G15400

MDP0000232309 MDP0000547450

chr16:13,766,655..13,768,223 chr7:26,032,235..26,033,783

BAX inhibitor-1 like protein UNC93-like protein 1-like

MDP0000712586 MDP0000161275 MDP0000310811

chr14:27,117,925..27,120,988 chr12:24,524,346..24,526,137 chr7:3,004,269..3,005,959

SCAI-like protein Mitochondrial succinate-fumarate transporter 1-like Cysteine proteinase inhibitor 6

[23,24]

Cell/Organ Development No hit AT1G18000 Stress Response No Hit AT5G01340 AT3G12490

Horticulturae 2016, 2, 11 Horticulturae 2016, 2, 11

9 of 17 9 of 17

Figure 1. GO-terms associated with excised and sequenced bands from the differential display Figure 1. GO-terms associated with excised and sequenced bands from the differential display procedure. Functional analysis of sequences was performed using the Blast2GO annotation suite procedure. Functional analysis of sequences was performed using the Blast2GO annotation suite from full length apple genome (v. 1.0) sequences to which differential display band sequences were from full length apple genome (v. 1.0) sequences to which differential display band sequences were matched via BLAST. matched via BLAST.

3.2. Quantitative Reverse Transcription PCR Validation of Differential Expression

3.2. Quantitative Reverse Transcription PCR Validation of Differential Expression

Differential expression observed in the differential display experiment was validated using

quantitative reverse-transcription PCRin (qPCR). This was performed all sampleswas whenvalidated at least oneusing Differential expression observed the differential display for experiment peel tissue exhibited differential PCR gene (qPCR). expression. A was second biological for replicate from the second quantitative reverse-transcription This performed all samples when at least growing season was also used to test for seasonal variation. Primer sets were designed based upon one peel tissue exhibited differential gene expression. A second biological replicate from the second predicted gene sequence custom script which selects 26-mer primers, and screens the entire growing season was also used to test for seasonal variation. Primer sets were designed based upon reference genome to ensure the specificity of the oligonucleotide. Primer sequences are provided in predicted gene sequence custom script which selects 26-mer primers, and screens the entire reference Table S3. Expression analysis was performed in triplicate for each gene along with a β-tubulin genome to ensure the specificity of the oligonucleotide. Primer sequences are provided in Table S3. control which exhibited a global standard deviation of